Type Alias AzureContentSafetyInput

Filter configuration for Azure Content Safety

type AzureContentSafetyInput = {
    hate?: AzureThreshold;
    prompt_shield?: boolean;
    self_harm?: AzureThreshold;
    sexual?: AzureThreshold;
    violence?: AzureThreshold;
}

Index

Properties

hate? prompt_shield? self_harm? sexual? violence?

Properties

`Optional`hate

hate?: AzureThreshold

`Optional`prompt_shield

prompt_shield?: boolean

Filter prompts for harmful content such as jailbreaks and prompt injections.

`Optional`self_harm

self_harm?: AzureThreshold

`Optional`sexual

sexual?: AzureThreshold

`Optional`violence

violence?: AzureThreshold